期刊文献+

面向移动终端的农业知识文语转换系统 被引量:2

Mobile Terminal-oriented Text to Speech System for Agriculture Knowledge
下载PDF
导出
摘要 设计了一套面向移动终端的农业知识文语转换系统,该系统借鉴语义检索技术中对于关键词的语义处理方式,用语义检索技术处理文本切分过程中所遇到的歧义字段,可将农业文本知识转换为自然流畅的语音,并结合呼叫中心技术、借助移动终端推广农业知识。首先对文语转换中的文本分析流程进行分析,明确歧义字段处理为文语转换的关键点。在明确关键点的基础上,设计出用于进行分词的词典,进而基于词典匹配和统计分析模型对歧义字段进行提取,再基于语义检索对歧义字段进行处理,从而实现歧义字段的切分,最终采用Cool Edit Pro 2.0软件实现语音合成和韵律处理功能,开发出面向移动终端的农业知识文语转换系统,可有效解决歧义字段的处理问题。由测试结果可以看出,本文算法的查准率为94.03%,较最大匹配法和三元语法的查准率分别提高了5.72个百分点和0.97个百分点;本文算法的查全率为95.32%,较最大匹配法和三元语法的查全率分别提高了0.23个百分点和1.95个百分点;本文算法的F-1测度为0.93,较最大匹配法提高了0.01,与三元语法相同,说明本文算法具有较好的性能。 With the help of semantic proce terminal-oriented text to speech system was ssing of keyword in semantic retrieval technology, a mobile designed for agriculture knowledge. The ambiguous phrase encountered in text segmentation were cutting by using semantic search technology which could converted agricultural text knowledge to smooth and natural voice, and popularized agricultural knowledge combining call center and mobile terminal technology. Firstly, the text analysis flow was analyzed, and the ambiguous phrase which was the key point of text to speech processing was confirmed. Based on this key point, the word segmentation dictionary was designed, and the ambiguous phrase was extracted based on the dictionary matching and statistical analysis. Secondly, the ambiguous phrase was processed based on semantic retrieval and the segmentation of ambiguous phrase was realized. Finally, the functions of speech synthesis and prosodic processing were realized with the software of Cool Edit Pro 2.0, and mobile terminal-oriented text to speech system was developed for agriculture knowledge which could effectively solve the problem of ambiguous phrase processing.
出处 《农业机械学报》 EI CAS CSCD 北大核心 2015年第1期266-271,239,共7页 Transactions of the Chinese Society for Agricultural Machinery
基金 山东省自主创新专项资助项目(2012CX90204)
关键词 农业知识 移动终端 文语转换 歧义字段 Agriculture knowledge Mobile terminal Text to speech Ambiguous phrase
  • 相关文献

参考文献10

  • 1闫引堂,周晓强.交集型歧义字段切分方法研究[J].情报学报,2000,19(6):637-643. 被引量:22
  • 2张彩琴,袁健.改进的正向最大匹配分词算法[J].计算机工程与设计,2010,31(11):2595-2597. 被引量:11
  • 3Niu Zhengyu,Chai Peigi. Segmentation of prosodic phrases for improving the naturalness of synthesized mandrine Chinese speech [ C ]//J Proceedings of ICSLP - 2000, 2000:39 - 45.
  • 4http: // www. moe. edu. cn/publicfiles/business/htmlfiles/moe/s230/201001/75598, html.
  • 5http :///projectileAs. Cs. emu. edu/research, publie/toolstsegmentationteval/index, html.
  • 6Schubert Foo, Hui Li. Chinese word segmentation and its effect on information retrieval [ J ]. Information Processing & Management, 2004,40( 1 ) : 161 - 190.
  • 7黄魏,高兵,刘异,杨克巍.基于词条组合的中文文本分词方法[J].科学技术与工程,2010,10(1):85-89. 被引量:6
  • 8王瑞雷,栾静,潘晓花,卢修配.一种改进的中文分词正向最大匹配算法[J].计算机应用与软件,2011,28(3):195-197. 被引量:41
  • 9Richard Tzong-Han Tsai. Chinese text segmentation: a hybrid approach using transductive learning and statistical association measures [ J]. Expert Systems with Applications,2010,37 (5) :3553 - 3560.
  • 10Fu Guohong, Kit Chunyu, Webster J J. Chinese word segmentation as morpheme-based lexical chunking [ J ]. Information Sciences, 2008,178 (9) : 2282 - 2296.

二级参考文献26

共引文献74

同被引文献8

引证文献2

二级引证文献5

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部